sorry for this late thing, i couldn't get the motivation to sit
down and fix this (some other stuff, like new passing of MTA args,
*sendmail-arguments* option etc., but..),

so hello again, Gavin,

Gavin Troy <[email protected]> wrote:
 |On Sat, Jun 14, 2014, at 16:42, Steffen Nurpmeso wrote:
 |> 
 |>|> # required syntax for < 14.7
 |>|> set [email protected]@mail.messagingengine.com=PASSWD
 |>|> # required syntax for 14.7
 |>|> set password-gavtroy%[email protected]=PASSWD
 |> 
 |> You are absolutely right, way too often URL percent encoded values
 |> are assumed when normal strings should be expected instead.
 |> There is a preliminary patch on [next].
 |
 |The "password-USER@HOST" still requires %40 in this preliminary patch.

I finally came up with this:

     Many variable options of S-nail exist in multiple versions: the plain
     `variable' as well as `variable-HOST' and `variable-USER@HOST'.  Here
     `HOST' indeed means `server:port' if a `port' had been specified in the
     respective URL, otherwise it refers to the plain `server'.  Also, `USER'
     isn't truly the `USER' that had been found when doing the user chain
     lookup as is described below, i.e., this `USER' will never be in URL per-
     cent encoded form, wether it came from an URL or not.

     E.g., wether an hypothetic URL `smtp://you%[email protected]' had been
     given that includes a user, or wether the URL was `smtp://our.house' and
     the user had been found differently, in order to lookup the variable
     smtp-use-starttls S-nail first looks for wether `smtp-use-starttls-
     you [email protected]' is defined, then wether `smtp-use-starttls-
     our.house' exists before finally ending up looking at the plain variable
     itself.

Could you please confirm this is also fixed for you?

 |>|> set password-mail.messagingengine.com:465=PASSWD
 |>|
 |>|which I think is not the desired behavior.

No, as above, i think this will remain a host/port tuple, just as
has been specified.

 |Actually, I don't know what the old/"correct" behaviour is, because I
 |never used pop/imap with a non-standard port! For smtp(s), the so-called
 |smtp-auth-user-USER@HOST and smtp-auth-password-USER@HOST settings were
 |used. Although strangely and misleadingly, "USER@HOST" has always
 |referred only to the value of "from", so a non-standard port never comes
 |into question.

S/MIME uses the value found in From: for signing and verification.
But i still think it makes sense to derive the rest of the user
credentials and anything else from the actual SMTP account.
I think for most people it doesn't make a difference, and there is
always the chance to get to that state by using `account's.
But it is definetely not nice.

--steffen

------------------------------------------------------------------------------
HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions
Find What Matters Most in Your Big Data with HPCC Systems
Open Source. Fast. Scalable. Simple. Ideal for Dirty Data.
Leverages Graph Analysis for Fast Processing & Easy Data Exploration
http://p.sf.net/sfu/hpccsystems
_______________________________________________
S-nail-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/s-nail-users

Reply via email to